The objective of this research is to develop a tissue magnetic susceptibility property analyzer for processing gradient echo (GRE) MRI data. GRE MRI has been routinely used in clinical practice. A major aspect of its image contrast is based on its unique sensitivity to tissue susceptibility, which demonstrates as hypointensive T2* contrast in the magnitude image. This susceptibility sensitivity is particularly useful for studying blood deoxyhemoglobin (foundation of fMRI) and blood breakdown products, methemoglobin, hemosiderin and ferritin (various bleeding disorders including traumatic brain injury, hemorrhage and microbleed, vascular disorders, neurodegenerative diseases, etc) that have strong susceptibilities. The visualization of T2* hypointensity contrasts, which can be further enhanced with a phase weighting as in SWI (susceptibility weighted imaging), are useful for detecting the presence of these bleeding disorders. However, T2* hypointensity contrasts contain substantial blooming artifacts that depend on imaging parameters (echo time, orientation, field strength, etc) and cause uncertainty in spatial localization. Furthermore, patient management may require assessing the disease severities, which demands careful data analysis, including quantification of high susceptibility sources. We propose to develop a novel morphology enabled dipole inversion (MEDI) approach for analyzing both phase and magnitude data in GRE MRI to extract tissue susceptibility. The phase images contain the magnetic field information for fitting susceptibility via Maxwell's Equation. The magnitude images contain tissue structure information for matching with susceptibility interfaces via least discordance. We have proved a mathematical theorem that these phase and magnitude information are sufficient to determine susceptibility. Assured by this theoretic existence proof, we propose to construct a robust and accurate inversion algorithm in this research. We have obtained very encouraging preliminary data indicating that our MEDI inverse approach is sufficiently accurate. !
This SBIR Phase I research project will develop a tissue magnetic susceptibility property analyzer for processing data from gradient echo (GRE) MRI that is routinely used in clinical practice. Quantitative mapping of tissue susceptibility is particularly useful for studying blood deoxyhemoglobin (foundation of fMRI) and blood breakdown products, methemoglobin, hemosiderin and ferritin (various bleeding disorders including traumatic brain injury, hemorrhage and microbleed, vascular disorders, neurodegenerative diseases, etc) that have strong susceptibilities. The successful outcome of this project will enable the dissemination of a standardized robust quantitative susceptibility mapping technology to the large MRI community. !
| Chang, Shixin; Zhang, Jingwei; Liu, Tian et al. (2016) Quantitative Susceptibility Mapping of Intracerebral Hemorrhages at Various Stages. J Magn Reson Imaging 44:420-5 |
| Wisnieff, Cynthia; Liu, Tian; Spincemaille, Pascal et al. (2013) Magnetic susceptibility anisotropy: cylindrical symmetry from macroscopically ordered anisotropic molecules and accuracy of MRI measurements using few orientations. Neuroimage 70:363-76 |
| Wang, Shuo; Lou, Min; Liu, Tian et al. (2013) Hematoma volume measurement in gradient echo MRI using quantitative susceptibility mapping. Stroke 44:2315-7 |
